The problem of Manx: Norse linguistic evidence for the survival of Manx Gaelic in the Scandinavian period
THE PROBLEM OF MANX
Norse linguistic evidence for the survival of Manx Gaelic in the Scandinavian period1
Abstract: This chapter is concerned with the question of whether Manx Gaelic continued to be spoken on the Isle of Man during the tenth to the thirteenth centuries, when the island was under Scandinavian rule. The inconclusive Gelling/Megaw debate in the late twentieth century, concerning the survival or extinction of Manx Gaelic during the Norse period, failed to account for linguistic evidence. This chapter considers a small amount of phonological and semantic evidence that lends support to the hypothesis that spoken Manx Gaelic survived during the Norse period.
This chapter examines the question of whether Manx Gaelic continued to be spoken on the Isle of Man under Scandinavian rule, during the tenth to the thirteenth centuries. The question is no longer sharply polarised as it was in the 1970s, with neither the survival nor extinction hypotheses now of major concern to Manx scholarship. Margaret Gelling arguing for extinction and Basil Megaw for survival used a body of evidence which is now somewhat exhausted, but the question remains unresolved.2 It is perhaps telling that, while the subject of debate was linguistic, neither of these two advocates was a historical linguist: Gelling was a toponymist, and Megaw an archaeologist, ethnologist, and museum director. Curiously, neither side used linguistic evidence as is discussed in this chapter, and which is used here to lend support to the survival hypothesis.
This chapter offers a linguistic perspective on the nature of cross-cultural interactions during the Norse period on the Isle of Man. Hitherto, little has been written concerning the ways in which language use might be relevant to understanding the social and cultural conditions in this period. In this chapter, a number of linguistic indicators of Manx Gaelic language survival are set against known historical conditions, and the potential for linguistic evidence to offer insight into historical society and culture is revealed.
Firstly, the Gelling/Megaw exchange is reviewed, and implications arising from the divergent methodological approaches that they used and ways in which sociolinguistic models might apply to historical contact between speakers of different languages are discussed. Following this, the chapter explores the linguistic evidence that might differentiate between the survival and extinction hypotheses, and finishes with observations on what this might mean for our understanding of Manx Gaelic, historically, in the context of both Norse influence and the other Gaelic languages.
Survival or extinction, and the limits to place-name evidence
In the 1970s, Gelling and Megaw came to represent the poles of a debate: did Manx Gaelic become extinct during Norse hegemony, or did it survive? Gelling, discussing earlier landmark Manx toponymic works of J. J. Kneen and Carl J. S. Marstrander, focused on thirteenth and fourteenth-century Manx place-names, as a basis for suggestions about the relative importance of Old Norse, Manx Gaelic, and English languages in those two centuries.5 Kneen’s view was that Gaelic in the Isle of Man was ‘replaced by a Scandinavian dialect’ and that ‘the earlier Gaelic population was either wiped out or absorbed’. In contrast, Marstrander argued that Manx bilingual stone crosses could be interpreted as refuting the view ‘that the Norwegians decimated or expelled the Gaelic population from the island’ instead arguing that there were ‘intimate and personal relations between the Gaidil and Gaill’, and that a Gaelic group coexisted with the Norwegians from the very beginning as ‘free and socially equal’.5
Gelling argues that between AD 900 and 1300, the Norse language predominated on Man and was more socially important. On the continued use of Gaelic, she used runic inscriptions and place-names to suggest a ‘relatively slight Gaelic linguistic survival during the period of Norse rule’/’ Paradoxically, she also notes the ‘virtual disappearance of Gaelic in the Norse period’.7 Gelling took the paucity of Gaelic place-names as evidence that the island’s population had not remained bilingual. While her place-name analyses may be sound, the methodology for extrapolating linguistic behaviour from surviving place-name records is wanting, failing to explore the critical relationship between inferred historical language use and how it might be possible to infer linguistic behaviours from surviving place-names. This is revealed, inter alia, through her imprecise terminology. She argues that the Irish names in the Manx runic corpus are ‘evidence of a Gaelic strain |... | in the population of Man, but not of the continuance of Gaelic speech’, but she does not define what a ‘strain’ is, and she seems to be using the appellative ‘Gaelic’ for non-linguistic characteristics.8 Similarly, her references to ‘slight linguistic survival’ and ‘virtual disappearance’ are not explained, and she is not clear what these might mean for linguistic behaviours. Is she talking about Gaelic spoken only in limited geographic areas, only used in narrowed or marginalised semantic fields (for example, for farming only), the decline or loss of referential, functional, or creative capacities in the language, a decline in numbers of Gaelic first-language speakers, or a decline over time in the transmission of Gaelic from generation to generation?9 She also does not negotiate the scope for bilingualism in the linguistic communities of medieval Norse Man. Is this ‘slight survival’ to be found in specific social classes, such as merchants, who are likely to have enjoyed interactions with other Gaelic-speaking areas of the British Isles?
Basil Megaw took issue with Gelling’s position in his 1976 article, reprinted with minor edits in 1978.10 Megaw charges Gelling with concluding that ‘Gaelic had ceased to be spoken in Man until reintroduced from outside the island (?Galloway) after 1300’, repeating his own position that Gaelic ‘held its own in the time of Manx kings of the Isles during the twelfth and thirteenth centuries’.11 Megaw introduces evidence for re-dating an ‘abbeyland bounds’ document to c. 1280 of the document, based on palaeographical analysis by Neil Ker,12 in which a particular entry reveals Gaelic place-name evidence dating to around 1257, thus arguing for Gaelic in use during the Scandinavian period. He also points to the absence of farm place-names in Gaelic achadh (‘enclosed field’) common in southwest Scotland and Ireland, arguing this presents ‘a real obstacle to the theory of any new Gaelic settlement in Man in the thirteenth century’, as re-settlement would bring with it the nomenclature common in the homelands of the settlers.13 Megaw emphasises the importance of using as wide a range of evidence as possible, and he includes epithets and patronymics (Gaelic forms datable to the Scandinavian period) alongside literary evidence (references in and to Irish and Gaelic texts during the period) and administrative history (possible Norse-period Gaelic origins of Manx administrative structures such as the Manx farm-land treeti). He ultimately argues that the conditions in Man were a balance between the powerful ‘Scandinavian element in the ruling circle and the chief land-owners’ and ‘a background of native continuity’ with widespread bilingualism in the population.14 However, Megaw’s lack of sound linguistic methodology', as with Gelling, is revealed in his imprecise terminology, where, for example, he talks of ‘two linguistic levels co-existing’ without explaining what these are.13
Megaw bristled at Margaret Gelling referring to part of his work where he attempts to establish the relative significance of Gaelic names among Norse in the documents as ‘unscholarly’, and Gelling herself makes her position clear in respect of this part, arguing that ‘it is unconvincing’.16 Megaw is emphatic about neither culturally nor linguistically compartmentalising Manx evidence, and he speculates about an already polyglot pre-Scandinavian Hebrides, suggesting that between the fifth and eighth centuries no fewer than four languages (British, Irish, English, and Latin) were spoken in Man. He concludes that ‘if we have learnt nothing else from this investigation, we now see that even our “Scandinavian” kings were characteristically Gaelic speakers’.17
In her 1978 reply, Gelling restates her argument that there is insufficient evidence to sustain Megaw’s view of Gaelic language survival through Norse occupancy of Man. She repeats her attribution to both Kneen and Marstrander ‘the view that the Gaelic language died out in Man for a period after the Norse settlement’, an attribution that is demonstrably wrong in respect of the latter, as noted above.18 Gelling happily embraces the re-dated bounds document, which Megaw depends upon, but disagrees broadly with his analysis of it, accepting that in the ‘district around The Curraghs is a limited area in which Gaelic speech either persisted from pre-Norse times, or was re-introduced at a comparatively early date’, but outside of that she remains in firm (and seemingly self-contradictory) disagreement with Megaw, that Gaelic died out, while admitting it might have persisted in that one area.19 Gelling restates her point that if Manx place-names are viewed chronologically then there was a period of Norse occupation when Gaelic stopped being used, with Norse the language of farmers and aristocracy alike. In the same volume, Danelaw place-name expert Gillian Fellows-Jensen sought some rapprochement between the two positions, endorsing Gelling’s arguments in favour of high Norse settlement density in Man, but also accepting Megaw’s possibility of continuity in spoken Gaelic.
The debate settled to a quiet détante after 1978, and a little more than a decade later Gelling softens her position, but still maintains the island was ‘predominantly (perhaps even exclusively) Norse-speaking’.20 In this later work, she repeated her mistaken declaration that Marstrander’s view was in support of the extinction model, and she gives very short shrift (just six lines) to Megaw’s opposing view.
Finally, in volume 3 of The New History of the Isle of Man, published in 2015, the topic is touched on by R. L. Thomson, but neither extinction nor survival is specifically endorsed." Thomson cites Gelling as agreeing with Marstrander (unreferenced by Thomson) that ‘Old Norse was the predominant language of Man c.900 to c.1300’, and quotes Kenneth Jackson to the effect that Man and the Hebrides were ‘united under Norse rule in a single Gaelic-speaking kingdom of the Isles until 1266’.22 Fellows-Jensen, in the same volume, also references Gelling when she reports that the low status of settlements bearing Gaelic place-names has been used ‘as an argument for a relatively slight survival of Gaelic names (and the Gaelic language) in the period of Viking rule and a subsequent réintroduction of the Gaelic language’.23
This debate exposed significant theoretical and methodological gaps between place-name studies and inferred language use, failing to problématisé whether or how place-names can be used as evidence for a community’s linguistic behaviours. Moreover, assumptions were made that were not explicitly identified or analysed. The polarisation seems best attributed to a kind of talking at cross-purpose, where common sense overlooks linguistics. Matters unresolved in the debate include the question of whether a place-name can stand as proxy for the language use of a community, and how that might take account of mixed and bilingual communities, and the corollary assumption that a lack of place-names in both languages can be used as evidence for a monolingual community.
Indeed, there is no necessary reason for a place-name to stand as verification of the language spoken, or even that the particular language endured through the period of that place-name’s existence. The establishment and recording of a Norse place-name cannot in itself preclude Gaelic being spoken at the same time, and given the inevitable flux in active communities, across a timeframe of three to four centuries, there can be no legitimate claim to static social, political, or cultural conditions. In 1972, Magne Oftedal explored an equivalent challenge in respect of Norse linguistic influence in Scottish Gaelic, in which he noted, inter alia, the separation between loanwords and place-names on a variety of methodological axes. While loanwords are understood as indicative of functional language use, place-name semantics are anchored to a location, and often to an historic moment. Similarly, in the same year, Alan Small questioned the idea that place-name densities on the Isle of Skye can be used to support theories of Viking settlement densities, while Charles W. J. Withers (1984) argued that Norse place-name density' in the Hebrides cannot be equated with either uniform or hegemonic linguistic usage.24
Moreover, the assumption that an absence of place-names of one sort is indicative of the absence of the people for whom such place-names hold meaning or importance belies the fact that there are always more people than there are place-names, and not all people will have a role in the naming of places. This is especially pertinent when considering the power differentials within a community, typically expressed in the control of both the written record and naming practices, coupled with the likely population influx and departure over time, whether Norse- or Gaelic-speaking.
In the absence of a sociolingüístic framework against which place-name evidence can be measured and analysed, there is no necessary historical causality' linking place-names with exclusivity of language use. The formal study' of toponyms has often been chiefly concerned with mapping genealogies and surface meanings of place-names, and with attributing explanatory power to the semantics. Botolv Helleland notes that ‘the etymological discussion and the historical bearings of place-names have traditionally attracted the widest interest among scholars’, but there has been a shift more recently towards a focus on ‘socio-onomastic and socio-psychological functions of place-names’.25 Considered in this light, interest in Manx Gaelic place-names has largely embraced just one aspect of the surface - the etymological and historical aspects of the apparent language of the name itself- and taken this as evidence that reflects on the communities living in those places over time. However, Helleland suggests that a ‘place name not only' points out a place, it also mediates a cluster of qualities and meanings attached to that place, partly valid for a single individual, partly shared by' a given social group’.26 Considering possible social and cognitive roles that place-naming can play, the Manx evidence would seem to have limited capacity for explicating the social conditions that predominated over a period of likely flux. For even if placenaming evidence exists in just one language, the absence of evidence to the contrary is certainly insufficient for a conclusion of absence of other language use. Moreover, underlying conditions of naming practice are likely' to be elusive.